Documentation
  • Getting Started
  • Creating a Product
    • Products Overview
    • Making a Helm Chart
      • Converting from docker-compose
      • Interpolated Values
      • Image Pull Secrets
      • Cluster Ingress
      • SSL Certificates
    • Setting up the Registry
      • Google Container Registry
  • Deploying a Product
    • Deployments Overview
    • Deploying on AWS
    • Deploying on GCP
    • Deploying on Azure (beta)
  • Debugging
    • GCP Permissions Errors
    • Volume Classes
Powered by GitBook
On this page
  • Deployment Architecture
  • How to deploy
  1. Deploying a Product

Deployments Overview

PreviousGoogle Container RegistryNextDeploying on AWS

Last updated 4 years ago

Deployment Architecture

Metamanagement deploys your vendor's services in as much isolation as we can. We isolate the infrastructure at a networking level, putting all services in their own VPC, but we also isolate the VPC itself from your main services, putting content in a sub-account or separate project to ensure that there's no ability for Metamanagement or your vendor to inappropriately access data or services that they should not.

This architecture decision affords us a few advantages. Because the customer owns the deployment in its entirety, they can change VPC ingress/egress settings, install endpoint monitoring solutions, and even set billing alerts to control how the infrastructure behaves.

Even more importantly, this architecture gives customers tight control over what data the newly installed service can access and what systems can access the newly installed service. By using either VPC peering or more selective cloud-specific networking primitives, customers can enjoy a tight flow of information above and beyond what they might expect in traditional on-premise settings.

How to deploy

For specific instructions on how to deploy, please consult the following pages:

Deploying on AWS
Deploying on GCP
Deploying on Azure (beta)